Case study - comparison of DE methods
Maria Doyle, Peter MacCallum Cancer Centre1
Stefano Mangiola, Walter and Eliza Hall Institute2
Source:vignettes/comparison_differential_transcription.Rmd
comparison_differential_transcription.Rmd
Introduction
tidybulk integrates several methods for differential transcript abundance testing.
Comparison of methods
We can serially perform differential analyses with several methods, and the results will be added to the original dataset.
We first pre-process the data, creating a tibble and identifying abundant genes.
pasilla_de <- rpharma2020tidytranscriptomics::pasilla %>% # Convert SE object to tibble tidybulk %>% # Scale abundance for plotting identify_abundant(factor_of_interest=condition)
This is an example for the default method for differential transcriptional testing. It uses the edgeR method.
pasilla_de %>% # Test differential composition test_differential_abundance( ~ condition + type, action="get" ) %>% arrange(FDR) #> tidybulk says: All methods use raw counts, #> irrespective of if scale_abundance or adjust_abundance have been calculated, #> therefore it is essential to add covariates such as batch effects (if applicable) in the formula. #> tidybulk says: The design column names are "(Intercept), conditionuntreated, typesingle_end" #> tidybulk says: to access the raw results (fitted GLM) do `attr(..., "internals")$edgeR` #> # A tibble: 14,599 x 7 #> feature .abundant logFC logCPM F PValue FDR #> <chr> <lgl> <dbl> <dbl> <dbl> <dbl> <dbl> #> 1 FBgn0025111 TRUE -2.86 6.92 1304. 2.37e-12 0.0000000149 #> 2 FBgn0039155 TRUE 4.61 5.88 1194. 3.75e-12 0.0000000149 #> 3 FBgn0003360 TRUE 3.12 8.45 792. 3.17e-11 0.0000000837 #> 4 FBgn0035085 TRUE 2.57 5.68 728. 4.89e-11 0.0000000968 #> 5 FBgn0029167 TRUE 2.19 8.22 599. 1.34e-10 0.000000169 #> 6 FBgn0034736 TRUE 3.52 4.18 586. 1.50e-10 0.000000169 #> 7 FBgn0039827 TRUE 4.17 4.39 616. 1.16e-10 0.000000169 #> 8 FBgn0026562 TRUE 2.47 11.8 504. 3.27e-10 0.000000323 #> 9 FBgn0000071 TRUE -2.63 4.79 389. 1.24e- 9 0.000000983 #> 10 FBgn0034434 TRUE 3.63 3.21 390. 1.22e- 9 0.000000983 #> # … with 14,589 more rows
Now let’s try to perform multiple methods on the same dataset.
de_all <- pasilla_de %>% # edgeR QLT test_differential_abundance( ~ condition + type, method = "edger_quasi_likelihood", prefix = "edgerQLT_" ) %>% # edgeR LRT test_differential_abundance( ~ condition + type, method = "edger_likelihood_ratio", prefix = "edgerLR_" ) %>% # limma-voom test_differential_abundance( ~ condition + type, method = "limma_voom", prefix = "voom_" ) %>% # DESeq2 test_differential_abundance( ~ condition + type, method = "deseq2", prefix = "deseq2_" )
We can visually compare the log fold change (logFC) of transcript abundance for the comparison of interest (treated vs untreated) for all methods. We will notice that the consistency of the logFC is really high for the methods.
de_all %>% keep_abundant() %>% pivot_transcript() %>% select(edgerQLT_logFC, edgerLR_logFC, voom_logFC, deseq2_log2FoldChange, feature ) %>% ggpairs(1:4)
Similarly, we can visually compare the significance for all methods. In this case the difference is larger.
de_all %>% keep_abundant() %>% pivot_transcript() %>% select(edgerQLT_PValue, edgerLR_PValue, voom_P.Value, deseq2_pvalue, feature ) %>% ggpairs(1:4)
We can investigate some of the genes that are less consistent between methods, for example between edgeR QLT and DESeq2. In this case we can use plotly to produce an interactive plot.
# Produce the ggplot ( de_all %>% keep_abundant() %>% ggplot(aes(edgerQLT_PValue, deseq2_pvalue, label = feature)) + geom_point() ) %>% # Make it interactive ggplotly()
We can select some of the transcripts for further analysis using the tidygate package.
de_gate = de_all %>% keep_abundant() %>% gate( feature, edgerQLT_PValue, deseq2_pvalue, opacity=0.3, how_many_gates = 2 )
#> # A tibble: 55,433 x 29
#> feature sample counts condition type .abundant edgerQLT_logFC
#> <chr> <fct> <int> <fct> <fct> <lgl> <dbl>
#> 1 FBgn00… untrt1 92 untreated sing… TRUE 0.0337
#> 2 FBgn00… untrt1 4664 untreated sing… TRUE 0.249
#> 3 FBgn00… untrt1 583 untreated sing… TRUE 0.0569
#> 4 FBgn00… untrt1 1446 untreated sing… TRUE 0.0402
#> 5 FBgn00… untrt1 15 untreated sing… TRUE -0.387
#> 6 FBgn00… untrt1 101664 untreated sing… TRUE -0.310
#> 7 FBgn00… untrt1 33402 untreated sing… TRUE -0.617
#> 8 FBgn00… untrt1 21 untreated sing… TRUE -0.491
#> 9 FBgn00… untrt1 12 untreated sing… TRUE -0.610
#> 10 FBgn00… untrt1 3026 untreated sing… TRUE 0.163
#> # … with 55,423 more rows, and 22 more variables: edgerQLT_logCPM <dbl>,
#> # edgerQLT_F <dbl>, edgerQLT_PValue <dbl>, edgerQLT_FDR <dbl>,
#> # edgerLR_logFC <dbl>, edgerLR_logCPM <dbl>, edgerLR_LR <dbl>,
#> # edgerLR_PValue <dbl>, edgerLR_FDR <dbl>, voom_logFC <dbl>,
#> # voom_AveExpr <dbl>, voom_t <dbl>, voom_P.Value <dbl>, voom_adj.P.Val <dbl>,
#> # voom_B <dbl>, deseq2_baseMean <dbl>, deseq2_log2FoldChange <dbl>,
#> # deseq2_lfcSE <dbl>, deseq2_stat <dbl>, deseq2_pvalue <dbl>,
#> # deseq2_padj <dbl>, gate <chr>
We can now select the transcripts from the two gates (i.e. oversignificant in DESeq2 and oversignificant in edgeR)
de_gate %>% scale_abundance() %>% # Filter only transcripts within the gates filter(gate > 0) %>% # Rename for clarity mutate(gate = case_when( gate == 1 ~ "more in edgeR", gate == 2 ~ "more in DESeq2", TRUE ~ gate )) %>% # Rearrange order mutate(feature = fct_reorder(feature, edgerQLT_PValue, min)) %>% # Plot ggplot(aes(condition, counts_scaled, color=gate)) + geom_point() + facet_wrap(~feature, scale="free_y") + custom_theme
For example DESeq2 produces a more conservative statistic for the transcript FBgn0052939.
Contributing
If you want to suggest improvements for this workshop or ask questions, you can do so as described here.
Reproducibility
Record package and version information with sessionInfo
sessionInfo() #> R version 4.0.2 Patched (2020-09-24 r79253) #> Platform: x86_64-pc-linux-gnu (64-bit) #> Running under: Ubuntu 20.04.1 LTS #> #> Matrix products: default #> BLAS/LAPACK: /usr/lib/x86_64-linux-gnu/openblas-pthread/libopenblasp-r0.3.8.so #> #> locale: #> [1] LC_CTYPE=en_US.UTF-8 LC_NUMERIC=C #> [3] LC_TIME=en_US.UTF-8 LC_COLLATE=en_US.UTF-8 #> [5] LC_MONETARY=en_US.UTF-8 LC_MESSAGES=C #> [7] LC_PAPER=en_US.UTF-8 LC_NAME=C #> [9] LC_ADDRESS=C LC_TELEPHONE=C #> [11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C #> #> attached base packages: #> [1] stats graphics grDevices utils datasets methods base #> #> other attached packages: #> [1] plotly_4.9.2.1 GGally_2.0.0 ggplot2_3.3.2 tidygate_0.2.8 tidybulk_1.1.8 #> [6] forcats_0.5.0 dplyr_1.0.2 #> #> loaded via a namespace (and not attached): #> [1] bitops_1.0-6 matrixStats_0.57.0 #> [3] fs_1.5.0 bit64_4.0.5 #> [5] RColorBrewer_1.1-2 httr_1.4.2 #> [7] rprojroot_1.3-2 GenomeInfoDb_1.25.11 #> [9] tools_4.0.2 backports_1.1.10 #> [11] utf8_1.1.4 R6_2.4.1 #> [13] DBI_1.1.0 lazyeval_0.2.2 #> [15] BiocGenerics_0.35.4 colorspace_1.4-1 #> [17] withr_2.3.0 tidyselect_1.1.0 #> [19] gridExtra_2.3 DESeq2_1.29.13 #> [21] bit_4.0.4 compiler_4.0.2 #> [23] preprocessCore_1.51.0 cli_2.0.2 #> [25] Biobase_2.49.1 desc_1.2.0 #> [27] DelayedArray_0.15.9 labeling_0.3 #> [29] scales_1.1.1 readr_1.3.1 #> [31] genefilter_1.71.0 pkgdown_1.6.1 #> [33] systemfonts_0.3.1 stringr_1.4.0 #> [35] digest_0.6.25 rmarkdown_2.3 #> [37] XVector_0.29.3 pkgconfig_2.0.3 #> [39] htmltools_0.5.0 limma_3.45.14 #> [41] htmlwidgets_1.5.1 rlang_0.4.7 #> [43] RSQLite_2.2.0 farver_2.0.3 #> [45] generics_0.0.2 rpharma2020tidytranscriptomics_1.0.6 #> [47] jsonlite_1.7.1 crosstalk_1.1.0.1 #> [49] BiocParallel_1.23.2 RCurl_1.98-1.2 #> [51] magrittr_1.5 GenomeInfoDbData_1.2.3 #> [53] Matrix_1.2-18 Rcpp_1.0.5 #> [55] munsell_0.5.0 S4Vectors_0.27.13 #> [57] fansi_0.4.1 viridis_0.5.1 #> [59] lifecycle_0.2.0 stringi_1.5.3 #> [61] yaml_2.2.1 edgeR_3.31.4 #> [63] SummarizedExperiment_1.19.6 zlibbioc_1.35.0 #> [65] plyr_1.8.6 blob_1.2.1 #> [67] grid_4.0.2 parallel_4.0.2 #> [69] crayon_1.3.4 lattice_0.20-41 #> [71] splines_4.0.2 annotate_1.67.1 #> [73] hms_0.5.3 locfit_1.5-9.4 #> [75] knitr_1.30 pillar_1.4.6 #> [77] GenomicRanges_1.41.6 geneplotter_1.67.0 #> [79] stats4_4.0.2 XML_3.99-0.5 #> [81] glue_1.4.2 evaluate_0.14 #> [83] data.table_1.13.0 BiocManager_1.30.10 #> [85] png_0.1-7 vctrs_0.3.4 #> [87] gtable_0.3.0 purrr_0.3.4 #> [89] tidyr_1.1.2 reshape_0.8.8 #> [91] assertthat_0.2.1 cpp11_0.2.1 #> [93] xfun_0.17 xtable_1.8-4 #> [95] survival_3.2-3 ragg_0.3.1 #> [97] viridisLite_0.3.0 tibble_3.0.3 #> [99] AnnotationDbi_1.51.3 memoise_1.1.0 #> [101] IRanges_2.23.10 ellipsis_0.3.1